We conducted a multicenter randomized double blinded controlled trial comparing early postnatal dexamethasone to selective late dexamethasone therapy in premature infants at 42 centers in the Vermont Oxford Network. Methods: Infants weighing 501-1000 grams were eligible for enrollment at 12 hours of age if they required assisted ventilation, had received surfactant replacement therapy, were physiologically stable, had no obvious life-threatening congenital anomaly, had blood cultures obtained and antibiotic therapy initiated. Infants were randomly assigned to early dexamethasone therapy or saline placebo. Intravenous dexamethasone was administered for 12 days according to the following dosing schedule: 0.5 mg/kg/day for three days, 0.25 mg/kg/day for three days, 0.10 mg/kg/day for three days, 0.05 mg/kg/day for three days. Infants in either group could receive treatment with late postnatal steroids beginning on day 14 of life if they were on assisted ventilation with supplemental oxygen greater than 30%. The primary outcome measure was chronic lung disease or death at 36 weeks adjusted age.